Before governments or school systems treat rising autism counts as evidence of a changing incidence and reallocate major resources, require a published robustness map that decomposes observed prevalence change into components (diagnostic substitution/accretion, registry/coverage changes, and residual incidence) using sibling controls, negative controls, E‑values and sensitivity bounds.
— Demanding standardized, auditable decompositions would prevent policy overreactions, target services where true need increased, and reduce politicized misinterpretation of administrative counts.
2009.10.04
100% relevant
King & Bearman’s 2009 California analysis estimated ~26% of the observed prevalence rise was due to diagnostic change via a documented pathway (MR→autism), illustrating the value of quantifying diagnostic‑practice effects before acting.
← Back to All Ideas